Search

Es un pasaje de lLas vestiduras recamadas de Salvador González Anaya se encuentran en animada charla dos de los protagonistas principales.

El surgimiento de la pintura en Málaga se debió a una serie de factores socieconómicos y culturales que fueron estudiados en mi libro Historia social de los pintores del siglo XIX en Málaga, al que remito al lector interesado en estos aspectos.

El artículo historia el desarrollo de las exposiciones de Bellas Artes y otras diversas durante los años que median entre 1843, en que se celebró la primera, y la de1862, que se hizo coincidir con la visita de S.M. La Reina a Málaga. Se pone en evidencia la distancia entre una modernidad que se desea y una sociedad todavía con ciertos aspec­tos propios de Antiguo Régimen.

We say that a Lorentzian metric and a semi-Riemannian metric on the same manifold M are null-projectively related if every null geodesic of the Lorentzian metric is an unparametrized geodesic of the semi-Riemannian one. This definition includes the case of conformally related Lorentzian metrics and the case of projectively equivalent metrics. We characterize the null-projectively relation by means of certain tensor and provide some examples. Then, we focus on the special case in which both metrics share parametrized null geodesics. In this case, it is said that they are null related. We show how to construct projectively equivalent metrics via a conformal transformation from null-related ones and conversely. The classical Levi-Civita theorem on projectively equivalent metrics is adapted to the case of null-related metrics and some results ensuring that two null-related metric are affinely equivalent are proven under curvature conditions.

By means of a counter-example, we show that the Reilly theorem for the upper bound of the first non-trivial eigenvalue of the Laplace operator of a compact submanifold of Euclidean space (Reilly, 1977, Comment. Mat. Helvetici, 52, 525–533) does not work for a (codimension ⩾2) compact spacelike submanifold of Lorentz–Minkowski spacetime. In the search of an alternative result, it should be noted that the original technique in (Reilly, 1977, Comment. Mat. Helvetici, 52, 525–533) is not applicable for a compact spacelike submanifold of Lorentz–Minkowski spacetime. In this paper, a new technique, based on an integral formula on a compact spacelike section of the light cone in Lorentz–Minkowski spacetime is developed. The technique is genuine in our setting, that is, it cannot be extended to another semi-Euclidean spaces of higher index. As a consequence, a family of upper bounds for the first eigenvalue of the Laplace operator of a compact spacelike submanifold of Lorentz–Minkowski spacetime is obtained. The equality for one of these inequalities is geometrically characterized. Indeed, the eigenvalue achieves one of these upper bounds if and only if the compact spacelike submanifold lies minimally in a hypersphere of certain spacelike hyperplane. On the way, the Reilly original result is reproved if a compact submanifold of a Euclidean space is naturally seen as a compact spacelike submanifold of Lorentz–Minkowski spacetime through a spacelike hyperplane.

In this paper, a new self-organizing artificial neural network called growing hierarchical neural forest (GHNF) is proposed. The GHNF is a hierarchical model based on the growing neural forest, which is a tree-based model that learns a set of trees (forest) instead of a general graph so that the forest can grow in size. This way, the GHNF faces three important limitations regarding the self-organizing map: fixed size, fixed topology, and lack of hierarchical representation for input data. Hence, the GHNF is especially amenable to datasets containing clusters where each cluster has a hierarchical structure since each tree of the GHNF forest can adapt to one of the clusters. Experimental results show the goodness of our proposal in terms of self-organization and clustering capabilities. In particular, it has been applied to text mining of tweets as a typical exploratory data analysis application, where a hierarchical representation of concepts present in tweets has been obtained. Moreover, it has been applied to foreground detection in video sequences, outperforming several methods specialized in foreground detection.

The design of automated video surveillance systems often involves the detection of agents which exhibit anomalous or dangerous behavior in the scene under analysis. Models aimed to enhance the video pattern recognition abilities of the system are commonly integrated in order to increase its performance. Deep learning neural networks are found among the most popular models employed for this purpose. Nevertheless, the large computational demands of deep networks mean that exhaustive scans of the full video frame make the system perform rather poorly in terms of execution speed when implemented on low cost devices, due to the excessive computational load generated by the examination of multiple image windows. This work presents a video surveillance system aimed to detect moving objects with abnormal behavior for a panoramic 360∘ surveillance camera. The block of the video frame to be analyzed is determined on the basis of a probabilistic mixture distribution comprised by two mixture components. The first component is a uniform distribution, which is in charge of a blind window selection, while the second component is a mixture of kernel distributions. The kernel distributions generate windows within the video frame in the vicinity of the areas where anomalies were previously found. This contributes to obtain candidate windows for analysis which are close to the most relevant regions of the video frame, according to the past recorded activity. A Raspberry Pi microcontroller based board is employed to implement the system. This enables the design and implementation of a system with a low cost, which is nevertheless capable of performing the video analysis with a high video frame processing rate.

The first author was partially supported by CAPES, Brazil. The second authors was partially supported by CNPq, Brazil, PQ-2 Grant 309668/2021-2. The third author was partially supported by Spanish MICINN Project PID2020-118452GB-I00. The fourth named author was partially supported by the Spanish MICINN and ERDF Project PID2020-116126GB-I00. The third and the fourth authors by the Andalusian and ERDF Project A-FQM-494-UGR18. Research partially supported by the ‘María de Maeztu’ Excellence Unit IMAG, reference CEX2020-001105-M, funded by MCIN-AEI-10.13039-501100011033., Coordenação de Aperfeiçoamento de Pessoal de Nível Superior CAPES, Conselho Nacional de Desenvolvimento Científico e Tecnológico 309668/2021-2 CNPq, Ministerio de Ciencia e Innovación PID2020-118452GB-I00 MICINN, European Regional Development Fund CEX2020-001105-M, MCIN-AEI-10.13039-501100011033, PID2020-116126GB-I00 ERDF

Pneumonia is an infectious and deadly disease which strikes over millions of people. Usually, chest X-rays are used by radiotherapist to diagnose pneumonia. In this paper, a Computer- Aided Diagnosis (CAD) system for pneumonia detection in chest X-ray images is proposed. This system is based on Convolutional Neural Networks (CNNs) which are able to classify the image into two classes (pneumonia or normal). Experimental results show that the proposed system obtained an accuracy rate of 98.59%. Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech.

New rigidity results for complete non-compact spacelike submanifolds of arbitrary codimension in plane fronted waves are obtained. Under appropriate assumptions, we prove that a complete spacelike submanifold in these spacetimes is contained in a characteristic lightlike hypersurface. Moreover, for a complete codimension two extremal submanifold in a plane fronted wave we show sufficient conditions to guarantee that it is a (totally geodesic) wavefront., Spanish Government PID2020-118452GBI00 PID2020-116126GB-I00, ERDF project P20-01391, "Maria de Maeztu" Excellence Unit IMAG - MCIN-AEI CEX2020-001105-M

For every codimension two spacelike submanifold of a Lorentz manifold and each choice of a normal lightlike vector field, we introduce a canonical way to construct a tractor conformal bundle. We characterize when the induced connection of a such submanifold defines a tractor connection and then, in this case, when this tractor conformal bundle with the induced connection is standard for the induced metric. Finally, the normality condition for this tractor conformal bundle endowed with the induced connection is characterized in terms of a strong relationship between the intrinsic and the extrinsic geometry of the starting spacelike submanifold., Comment: 10 pages

Objective To determine if the two-minute walk test (2MWT) could serve as an alternative measure of high-level mobility in lower limb prosthesis users when circumstances preclude administration of the Comprehensive High-level Activity Mobility Predictor (CHAMP). Design Cross-sectional study. Setting Indoor recreational athletic field and gymnasium Subjects Fifty-eight adult lower limb prosthesis users with unilateral or bilateral lower limb amputation who participate in recreational athletic activities. Intervention N/A Main Measures The 2MWT and CHAMP while using their preferred prosthesis(es) on an indoor artificial athletic field or hardwood gymnasium floor. Results Thirty-nine men and nineteen women with a median age of 38.3 years participated in the study. Most participants experienced amputation(s) due to trauma (62%) or tumor (10%) and were generally higher functioning (K4 (91.4%) and K3 (8.6%)). The median (range) score for the CHAMP was 23.0 points (1.5–33.5) and the mean ± standard deviation (range) 2MWT distance walked was 188.6 ± 33.9 m (100.2–254.3 m). The CHAMP demonstrated a strong positive relationship with 2MWT (r = 0.83, p Conclusions Although the 2MWT does not test multi-directional agility like the CHAMP, they were found to be highly correlated. If space is limited, the two-minute walk test can serve as an alternative measure for assessing high-level mobility capabilities in lower limb prosthesis users.

To analyse the COVID-19-related lockdown impact on University workers, to identify groups based on this information, and to study the factors associated with each group.Cross-sectional study.A survey was conducted 3.5 weeks after COVID-19-related lockdown in University workers in Spain. Sociodemographic variables, housing, work, health conditions, levels of anxiety, stress and depression (DASS-21), and social support (MSPSS) were collected. A cluster analysis was performed to identify groups depending on the impact of the lockdown. Differences between groups were tested using Chi-square and Mann-Whitney-U tests, and associated factors with binary logistic regression.We identified two groups of workers. "G1: Consequences in the daily life routine" was mainly composed of men, Research and Teaching Personnel (RTP) with more stable professional categories, higher income level, and bigger houses than people in G2. Participants in "G2: Concerns for the current and future well-being" presented worse intensity of pain than before the lockdown, more anxiety, depression, stress and less social support than people in G1. ASP (Administration and Services Personnel) had more risk of belonging to G2 than RTP (OR = 5.863). A higher number of people living at home decreased the risk of being in G2 (OR = 0.439). People with lower pain intensity had less risk of being in G2 (OR = 0.014), and this risk decreased as friends support increased (OR = 0.833).In G1, the consequences were immediately reflected in the stress resulting from changes in their daily work routine. In G2, the concerns were related to their professional future, with worse mental health, greater intensity of pain and less social support.

Introduction Subclinical atherosclerosis improves cardiovascular risk assessment and is considered a risk modifier in individuals at borderline-to-intermediate estimated risk by clinical risk scores. Purpose We seek to propose a new decision algorithm for statin allocation based on the quantification of carotid and/or femoral atherosclerosis burden by novel 3-dimensional vascular ultrasound (3DVUS) and to explore its potential additive value when used in combination with conventional risk evaluation. Methods We conducted an observational study in all outpatients without previous history of cardiovascular events assessed in the Cardiovascular Disease Prevention Unit during 2017–2020. Cardiovascular risk was assessed with the ACC/AHA 10-year atherosclerotic cardiovascular disease (ASCVD) risk algorithm to establish the indication for initiating statin therapy following current clinical guidelines. All patients underwent 3DVUS evaluation for carotid plaque burden (CPB) calculated as the sum of all plaque volumes present in bilateral carotid arteries. Femoral 3DVUS evaluation was included later in the study protocol. Global plaque burden (GPB) was calculated in patients with both carotid and femoral 3DVUS as the sum of all plaque volumes present in bilateral carotid and femoral arteries. Carotid and global plaque burden was classified as high, moderate and low if a patient presented a percentile (p) >75, p50–75 and p75) or discouraging (p Results One hundred sixty-three patients (age 51±8 years; 72,4% men; 10y-ASCVD 5,9%±4,8%) were included, being the most frequent reason for referral the presence of metabolic syndrome (37%), followed by intermediate or indeterminate estimated CV risk (33%), one markedly elevated single risk factor (15%) and family history of early CVD (13%). 10y-ASCVD score classified 80 (49%) patients as low-risk without recommendation for initiating statins, 18 (11%) as high-risk indicating statin therapy and 65 (40%) patients that needed clinical-patient risk discussion. Among patients under risk discussion, the CPB percentile re-stratified 51 (78%) of them, recommending statins in 26 (40%) and discouraging statins in 25 (38%). In addition, CPB re-stratified 24 (30%) low-risk patients to recommend statins. A sub-group of 114 patients additionally underwent femoral 3DVUS, and calculated GPB led to similar results of patient re-stratification (Figure). Conclusions Quantification of 3D atherosclerosis burden and its percentile would re-stratify a significant number of patients with intermediate risk, becoming a potentially useful tool for clinical decision making. In addition, it possibly improves the detection of low-risk patients who would benefit from statin therapy. Funding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Dr. Lόpez-Melgar has received a grant from the Spanish Society of Cardiology “Proyecto de investigaciόn traslacional en Cardiología 2020” Patient re-stratification with CPB/GPB

In this paper, a new color quantization method based on a self-organized artificial neural network called the Growing Hierarchical Bregman Neural Gas (GHBNG) is proposed. This neural network is based on Bregman divergences, from which the squared Euclidean distance is a particular case. Thus, the best suitable Bregman divergence for color quantization can be selected according to the input data. Moreover, the GHBNG yields a tree-structured model that represents the input data so that a hierarchical color quantization can be obtained, where each layer of the hierarchy contains a different color quantization of the original image. Experimental results confirm the color quantization capabilities of this approach.

Social conflicts appearing in the media are increasing public awareness about security issues, resulting in a higher demand of more exhaustive environment monitoring methods. Automatic video surveillance systems are a powerful assistance to public and private security agents. Since the arrival of deep learning, object detection and classification systems have experienced a large improvement in both accuracy and versatility. However, deep learning-based object detection and classification systems often require expensive GPU-based hardware to work properly. This paper presents a novel deep learning-based foreground anomalous object detection system for video streams supplied by panoramic cameras, specially designed to build power efficient video surveillance systems. The system optimises the process of searching for anomalous objects through a new potential detection generator managed by three different multivariant homoscedastic distributions. Experimental results obtained after its deployment in a Jetson TX2 board attest the good performance of the system, postulating it as a solvent approach to power saving video surveillance systems.

The space of invariant affine connections on every 3-Sasakian homogeneous manifold of dimension at least seven is described. In particular, the subspace of invariant affine metric connections and the subclass with skew torsion are also determined. To this aim, an explicit construction of all 3-Sasakian homogeneous manifolds is exhibited. It is shown that the 3-Sasakian homogeneous manifolds which admit nontrivial Einstein with skew torsion invariant affine connections are those of dimension seven, that is, $${\mathbb {S}}^7$$, $${\mathbb {R}}P^7$$ and the Aloff–Wallach space $${\mathfrak {W}}^{7}_{1,1}$$. On $${\mathbb {S}}^7$$ and $${\mathbb {R}}P^7$$, the set of such connections is bijective to two copies of the conformal linear transformation group of the Euclidean space, while it is strictly bigger on $${\mathfrak {W}}^{7}_{1,1}$$. The set of invariant connections with skew torsion whose Ricci tensor satisfies that its eigenspaces are the canonical vertical and horizontal distributions, is fully described on 3-Sasakian homogeneous manifolds. An affine connection satisfying these conditions is distinguished, by parallelizing all the Reeb vector fields associated with the 3-Sasakian structure, which is also Einstein with skew torsion on the 7-dimensional examples. The invariant metric affine connections on 3-Sasakian homogeneous manifolds with parallel skew torsion have been found. Finally, some results have been adapted to the non-homogeneous setting.

In this work we introduce a variant of the well-known Traveling Purchaser Problem in which the purchased products must be delivered to a set of customers. The objective is to minimize the total cost (purchasing plus traveling costs) and the waiting time of the customers, simultaneously, while satisfying the total demand. This problem is called the bi-objective Traveling Purchaser Problem with Deliveries. In order to approximate Pareto fronts for this problem, a relinked variable neighborhood search is proposed and tested over a large set of artificial instances. Our results show that our algorithm is highly competitive compared to the ϵ-constraint method in small instances. On the other hand, experiments carried out over large instances show that our algorithm is able to find Pareto front approximations with more points in a shorter running time for uncapacitated instances than for capacitated ones. Also, computational results show that the performance of some local searches used in our algorithm depends on the characteristics of the instances, this underlines the importance of designing a metaheuristic based on multiple local searches.

In view of the extensive use of hexagonal boron nitride (hBN) in 2D material electronics, it becomes important to refine its dielectric characterization in terms of low-field permittivity and high-field strength and conductivity up to the breakdown voltage. The present study aims at filling this gap using DC and RF transport in two Au-hBN-Au capacitor series of variable thickness in the 10--100 nm range, made of large high-pressure, high-temperature (HPHT) crystals and a polymer derivative ceramics (PDC) crystals. We deduce an out-of-plane low field dielectric constant $\epsilon_\parallel=3.4\pm0.2$ consistent with the theoretical prediction of Ohba et al., that narrows down the generally accepted window $\epsilon_\parallel=3$--$4$. The DC-current leakage at high-field is found to obey the Frenkel-Pool law for thermally-activated trap-assisted electron transport with a dynamic dielectric constant $\epsilon_\parallel\simeq3.1$ and a trap energy $\Phi_B\simeq1.3\;\mathrm{eV}$, that is comparable with standard technologically relevant dielectrics., Comment: 10 pages, 2 figures, AM version in Materials Research Express

The invariant metric affine connections on Berger spheres which are Einstein with skew-torsion are determined in both Riemannian and Lorentzian signature. Expressions of such connections are explicitly given. Notably, every Berger sphere with Lorentzian signature admits invariant metric affine connections which are Einstein with skew-torsion except for dimension 3. For Riemannian signature, the existence of such connections strongly depends on the dimension of the sphere and on the scale of the deformation used for the Berger metric. In particular, there are Riemannian Berger spheres, not Einstein, which admit invariant Einstein with skew-torsion affine connections.

Lightlike Cartan geometries are introduced as Cartan geometries modelled on the future lightlike cone in Lorentz-Minkowski spacetime. Then, we provide an approach to the study of lightlike manifolds from this point of view. It is stated that every lightlike Cartan geometry on a manifold $N$ provides a lightlike metric $h$ with radical distribution globally spanned by a vector field $Z$. For lightlike hypersurfaces of a Lorentz manifold, we give the condition that characterizes when the pull-back of the Levi-Civita connection form of the ambient manifold is a lightlike Cartan connection on such hypersurface. In the particular case that a lightlike hypersurface is properly totally umbilical, this construction essentially returns the original lightlike metric. From the intrinsic point of view, starting from a given lightlike manifold $(N,h)$, we show a method to construct a family of ambient Lorentzian manifolds that realize $(N,h)$ as a hypersurface. This method is inspired on the Feffermann-Graham ambient metric construction in conformal geometry and provides a lightlike Cartan geometry on the original manifold when $(N,h)$ is generic.

We analyse the intrinsic and extrinsic geometry of spacelike submanifolds in light cones $$\Lambda _{c}(p)$$ of de Sitter and anti-de Sitter spacetimes. Every light cone $$\Lambda _{c}(p)$$ contains the lightlike geodesics starting from p and, essentially, it coincides with the horisms $$E^{+}(p)\cup E^{-}(p)$$ . The analysis works by means of an explicit correspondence with the spacelike submanifolds through the light cone in the Lorentz–Minkowski spacetime. In particular, a characterization of totally umbilical compact surfaces through light cones in de Sitter and anti-de Sitter is shown and we obtain an estimation of the first eigenvalue of the Laplace operator on a compact spacelike surface in a light cone.

En este trabajo se describen los componentes y ejes de actuación del Plan de Recuperación, Transformación y Resiliencia (PRTR) y el posible destino de los fondos que comprende, con especial incidencia en sectores particularmente significativos en el ámbito de la economía social y de las empresas de participación, como es el caso del medio agropecuario. Precisamente, se valora la oportunidad de utilización de estos fondos para proyectos vinculados al medio rural y con un vector de digitalización preponderante. La adopción de tecnologías como la inteligencia artificial, la conectividad IoT, la computación en la nube o las propiedades de trazabilidad de blockchain pueden constituir un revulsivo para afrontar, también, una evolución, transición, e incluso, una revolución hacia un entorno “Agro 4.0” en similitud con la actual conceptualización de la denominada Industria 4.0. Aunque los fondos europeos no distinguen ni parcelan entre formas jurídicas; sí procede analizar los fondos destinados a algunos de los sectores donde la economía social tiene mayor representatividad en España y, particularmente, el caso del sector agrícola por sus múltiples implicaciones de carácter económico-social, como la cuestión de la despoblación (la España Vaciada), la preservación medioambiental, la fijación de empleo en el medio rural y otras áreas derivadas.

To evaluate whether Patient-Reported Outcomes Measurement Information System (PROMIS) pediatric patient-reported outcome (PRO) measures can serve as valid endpoints in a clinical trial of a chronic pediatric illness.We evaluated the responsiveness of PROMIS pediatric measures collected through the Clinical Outcomes of Methotrexate Binary Therapy in Practice (COMBINE) trial, a multicenter, randomized, double-blind, placebo-controlled, pragmatic clinical trial in pediatric patients with Crohn's disease (CD). We examined the relationships between changes in PROMIS pediatric measures and changes in disease activity by evaluating PRO score changes among patients who did and patients who did not experience improvement in disease activity.Participants included 266 children and adolescents with CD from a total of 35 institutions. Over the course of follow-up, participants showed improvement in most PRO domains, with the largest effect sizes observed for the clinically improved group. Patients who maintained steroid-free remission showed significantly lower PRO scores for the Pain Interference, Fatigue, and inflammatory bowel disease (IBD) Symptoms domains and higher scores for the Positive Affect domain.This study demonstrates the responsiveness of the PROMIS pediatric measures of Fatigue and Pain Interference as study endpoints in a large, multicenter pragmatic trial in pediatric CD, extending a growing body of research supporting the use of PROMIS pediatric measures as reliable PRO endpoints for clinical trials.

Clustering is one of the most common applications of unsupervised learning, being present in many statistical data analysis processes performed by scientists and engineers. Because of their special features, some categories of Artificial Neural Networks have demonstrated to be specially efficient when it comes to clustering. The Growing Neural Gas (GNG) is a good example of these networks, not only because its capability for revealing the clusters underlying in a certain distribution with an optimized number of neurons, but to faithfully describe the topological relations among the different clusters of a dataset. However, because of their intrinsic nature, there will be some data distributions with regions where no data can be found. Aiming to perform a clustering process on these datasets, this paper presents the design of a Growing Neural Gas-inspired model that keeps its neuron prototypes out of a set of regions previously specified, namely Forbidden Region Growing Neural Gas (FRGNG). Experimental results illustrate how this model can represent an alternative, in terms of accuracy, to one of the most recent region avoiding clustering algorithms such as the Forbidden Region Self-Organizing Map (FRSOFM).

Most clustering algorithms are designed to minimize a distortion measure which quantifies how far the elements of the clusters are from their respective centroids. The assessment of the results is often carried out with the help of cluster quality measures which take into account the compactness and separation of the clusters. However, these measures are not amenable to optimization because they are not differentiable with respect to the centroids even for a given set of clusters. Here we propose a differentiable cluster quality measure, and an associated clustering algorithm to optimize it. It turns out that the standard k-means algorithm is a special case of our method. Experimental results are reported with both synthetic and real datasets, which demonstrate the performance of our approach with respect to several standard quantitative measures.

By means of several counterexamples, the impossibility to obtain an analogue of the Chen lower estimation for the total mean curvature of any compact submanifold in Euclidean space for the case of compact space-like submanifolds in Lorentz–Minkowski spacetime is shown. However, a lower estimation for the total mean curvature of a four-dimensional compact space-like submanifold that factors through the light cone of six-dimensional Lorentz–Minkowski spacetime is proved by using a technique completely different from Chen's original one. Moreover, the equality characterizes the totally umbilical four-dimensional round spheres in Lorentz–Minkowski spacetime. Finally, three applications are given. Among them, an extrinsic upper bound for the first non-trivial eigenvalue of the Laplacian of the induced metric on a four-dimensional compact space-like submanifold that factors through the light cone is proved.

In this paper, we develop a formula for spacelike surfaces in a four-dimensional Lorentzian space form which involves its mean curvature vector field, the Gauss curvature of the induced metric and the Gauss curvature of the second fundamental form associated to a non-degenerate null normal section. By means of this formula, we establish several sufficient conditions for a compact spacelike surface in a four-dimensional Lorentzian space form which has a null umbilical normal direction. As another application, we give a new proof of Liebmann rigidity theorems in Euclidean, hemispherical, hyperbolic spaces and in the De Sitter spacetime.

We model and solve a real-life distribution problem faced by a fresh fruit supplier. This problem is formulated as a Multi-Product Split Delivery Capacitated Team Orienteering Problem with Incomplete Service and Soft Time Windows. The problem is modeled through a mixed integer linear programming formulation and solved by an Adaptive Large Neighborhood Search (ALNS) metaheuristic. Computational results over a large set of artificial instances show that the combination of ALNS with a multi-start scheme produces better results than a classical implementation of the ALNS in which a single solution is built and improved.

This paper addresses a variant of the Orienteering Problem taking into account mandatory visits and exclusionary constraints (conflicts among nodes). Five mixed integer linear formulations are adapted from the Traveling Salesman Problem literature in order to provide a robust formulation for this problem. The main difference among these formulations lies in the way they deal with the subtour elimination constraints. The performance of the proposed formulations is evaluated over a large set of instances. Computational results reveal that the model that avoids subtours by means of a single-commodity flow formulation allows to solve to optimality more instances than the other formulations, within a time limit of 1 h.

This paper addresses a variant of the Orienteering Problem in which some constraints related to mandatory visits and incompatibilities among nodes are taken into account. A hybrid algorithm based on a reactive GRASP and a general VNS is proposed. Computational experiments over a large set of instances show the efficiency of the algorithm. Additionally, we also validate the performance of this algorithm on some instances taken from the literature of the traditional Orienteering Problem. HighlightsThis paper addresses the Orienteering Problem (OP) with mandatory visits and exclusionary constraints.An efficient procedure based on a reactive GRASP and a general VNS is proposed.Extensive computational results on artificial instances are analyzed.Additionally, the performance of the proposed procedure is validated on a set of instances taken from the literature of the traditional OP.

A low cost smart motion detector is presented.It is based on the Arduino DUE microcontroller.The software architecture employs a fixed point arithmetic paradigm.The self-organizing map neural network is implemented on chip.The performance is substantially higher than that of the traditional detector. Most current approaches to computer vision are based on expensive, high performance hardware to meet the heavy computational requirements of the employed algorithms. These system architectures are severely limited in their practical application due to financial and technical limitations. In this work a different strategy is used, namely the development of an inexpensive and easy to deploy computer vision system for motion detection. This is achieved by three means. First of all, an affordable and flexible hardware platform is employed. Secondly, the motion detection algorithm is specifically tailored to involve a very small computational load. Thirdly, a fixed point programming paradigm is followed in implementing the system so as to further reduce the computational requirements. The proposed system is experimentally compared to the standard motion detector for a wide range of benchmark videos. The reported results indicate that our proposal attains substantially better performance, while it remains affordable and easy to install in practice.

Self-organizing maps (SOMs) are aimed to learn a representation of the input distribution which faithfully describes the topological relations among the clusters of the distribution. For some data sets and applications, it is known beforehand that some regions of the input space cannot contain any samples. Those are known as forbidden regions. In these cases, any prototype which lies in a forbidden region is meaningless. However, previous self-organizing models do not address this problem. In this paper, we propose a new SOM model which is guaranteed to keep all prototypes out of a set of prespecified forbidden regions. Experimental results are reported, which show that our proposal outperforms the SOM both in terms of vector quantization error and quality of the learned topological maps.

Automatic video surveillance systems are usually designed to detect anomalous objects being present in a scene or behaving dangerously. In order to perform adequately, they must incorporate models able to achieve accurate pattern recognition in an image, and deep learning neural networks excel at this task. However, exhaustive scan of the full image results in multiple image blocks or windows to analyze, which could make the time performance of the system very poor when implemented on low cost devices. This paper presents a system which attempts to detect abnormal moving objects within an area covered by a 360\(^\circ \) camera. The decision about the block of the image to analyze is based on a mixture distribution composed of two components: a uniform probability distribution, which represents a blind random selection, and a mixture of Gaussian probability distributions. Gaussian distributions represent windows in the image where anomalous objects were detected previously and contribute to generate the next window to analyze close to those windows of interest. The system is implemented on a Raspberry Pi microcontroller-based board, which enables the design and implementation of a low-cost monitoring system that is able to perform image processing.

Electronic noses are sensing devices that are able to classify chemical volatiles according to the readings of an array of non-selective gas sensors and some pattern recognition algorithm. Given their high versatility to host multiple sensors while still being compact and lightweight, e-noses have demonstrated to be a promising technology to real-world chemical recognition, which is our main concern in this work. Under these scenarios, classification is usually carried out on sub-sequences of the main e-nose data stream after a segmentation phase which objective is to exploit the temporal correlation of the e-nose's data. In this work we analyze to which extent considering segments of delayed samples by means of fixed-length sliding windows improves the classification accuracy. Extensive experimentation over a variety of experimental scenarios and gas sensor types, together with the analysis of the classification accuracy of three state-of-the-art classifiers, support our conclusions and findings. In particular, it has been found that fixed-length sliding windows attain better results than instantaneous sensor values for several classifier models, with a high statistical significance.

Background segmentation methods are exposed to the effects of different kinds of noise due to the limitations of image acquisition devices. This type of distortion can worsen the performance of segmentation methods because the input pixel values are altered. In this paper we study how several well-known background segmentation methods perform when the input is corrupted with several levels of uniform and Gaussian noise. Furthermore, few situations are reported where instead of an inconvenience, adding noise to the input may be desirable to attenuate some limitations of a method. In this work, the performance of nine well known methods is studied under both kinds of noise.

The discovery of the underlying topology of real-world data is a difficult task due to the high-dimensional and the complex structure in real datasets. In some cases, when the topology of the data is not known or the information is provided in a stream, it is advantageous to learn tree topologies from the data. This task can be carried out by dynamic self-organizing neural networks, so that the specific topology of the dataset is discovered by the network. In this work a self-organizing spanning tree is proposed, which is able to learn a tree topology without any prespecified structure. Experimental results are provided to show the performance of the model with real video data for a foreground detection task. Comparative results are reported.

In this paper, a new self-organizing neural gas model that we call Growing Hierarchical Bregman Neural Gas (GHBNG) has been proposed. Our proposal is based on the Growing Hierarchical Neural Gas (GHNG) in which Bregman divergences are incorporated in order to compute the winning neuron. This model has been applied to anomaly detection in video sequences together with a Faster R-CNN as an object detector module. Experimental results not only confirm the effectiveness of the GHBNG for the detection of anomalous object in video sequences but also its selforganization capabilities. Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech

The Iberian Peninsula is a major European region of biodiversity, as it harbours more than 30% of European endemic species. Despite a number of studies having evaluated the ability of nature reserves to protect certain taxa, there is still a lack of knowledge on how Iberian endemic fauna are represented in these reserves. We detected biodiversity hotspots of Iberian endemicity and evaluated the effectiveness of the Natura 2000 network (N2000) in representing 249 endemic species from eight animal taxonomic groups (amphibians, mammals, freshwater fishes, reptiles, water beetles, butterflies, lacewings and dung beetles). We found that only the 10% of these Iberian endemic species are considered species of community interest (i.e. species included in the Annexes of the Habitats Directive). We conducted gap analyses and null models of representativeness in N2000. Generally, N2000 is effective in its representation of Iberian endemic fauna, although we detected species and few hotspots of endemism that were still not represented. It is necessary to declare a few new protected areas, thus enhancing N2000's effectiveness in the conservation of the Iberian endemic fauna. Although the aim of N2000 is to protect species listed in the Birds and Habitats Directives, the conservation status of endemic species from one of the most important areas of Europe in terms of biodiversity, could be also a concern for the European Union. Our results are useful in the context of the recent European Commission mandate calling for a ‘fitness check’ of the Birds and Habitats Directives. This approach could be also applicable to other regions with high value of endemicity., DS‐F was supported by a post‐doctoral contract funded by Universidad de Castilla‐La Mancha and the European Social Fund (ESF). PA was supported by a ‘Ramón y Cajal’ contract (RYC‐2011‐07670, MINECO). This research was partially funded by project POII11‐0277‐5747 (Junta de Castilla‐La Mancha).

Automatic video surveillance systems are usually designed to detect anomalous objects being present in a scene or behaving dangerously. In order to perform adequately, they must incorporate models able to achieve accurate pattern recognition in an image, and deep learning neural networks excel at this task. However, exhaustive scan of the full image results in multiple image blocks or windows to analyze, which could make the time performance of the system very poor when implemented on low cost devices. This paper presents a system which attempts to detect abnormal moving objects within an area covered by a PTZ camera while it is panning. The decision about the block of the image to analyze is based on a mixture distribution composed of two components: a uniform probability distribution, which represents a blind random selection, and a mixture of Gaussian probability distributions. Gaussian distributions represent windows in the image where anomalous objects were detected previously and contribute to generate the next window to analyze close to those windows of interest. The system is implemented on a Raspberry Pi microcontroller-based board, which enables the design and implementation of a low-cost monitoring system that is able to perform image processing. Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech.